Help me understand the QRV outputs on my 266r

Discussing some incredible modules that don't quite fit into the other forum categories.

Moderators: Joe., lisa, luketeaford, Kent

Post Reply
User avatar
njdoyle
Common Wiggler
Posts: 171
Joined: Tue Jan 28, 2014 5:00 pm
Location: Ottawa
Contact:

Help me understand the QRV outputs on my 266r

Post by njdoyle » Sun Jul 23, 2017 6:21 pm

I just got a 266r and am very much enjoying it.

Everything behaves as I expect except for the quantized random voltages section. I don't believe my module is broken or defective or anything because the behaviour I'm getting matches the behaviour of demos of the Sputnik West Coast Random Source I've seen. I do wonder, though, if the implementation of the "r" QRV matches the behaviour on a Buchla 266.

I've noticed two things.

The first thing I noticed is that I'm convinced the 2^n and the n+1 outputs are reversed from what they're supposed to be. When I set n=6 (full) I definitely only get seven discreet values out of the top (2^n) output and I get lots of different discreet values out of the bottom (n+1) output.

The second thing I noticed is that for all values of n, I get outputs that span pretty much the entire CV range. For example, at n=1 I get very low or very high output from both outputs. As I increase n, the space between gets filled in. This is different from what I expected. I expected the upper end of the output level to increase with n. The way I understood it worked was that the output that gives n+1 outputs 1.2V per n (octaves). Higher n means higher possible octaves. The same kind of thing for the 2^n output but in 0.1V increments (semitones). Higher n means higher possible semitones.

For comparison, the Doepfer version behaves like I expect (different from my 266r). This video demonstrates:

Can anyone tell me if any of this makes sense or help me understand what's going on? Does anyone know if the "r" and the Buchla versions differ or is it my understanding that's missing something?

User avatar
davebr
Veteran Wiggler
Posts: 668
Joined: Mon Sep 07, 2009 12:55 am
Location: Portland, OR USA

266 QRV

Post by davebr » Mon Jul 24, 2017 10:36 am

I have a description with scope images on my 5U 266 module page and some more on my Buchla 266 module pages. Your steps are always the full CV range. N+1 simply divides the CV range by the number of steps. 2^N is binary weighted.

Dave

User avatar
addendum
Common Wiggler
Posts: 127
Joined: Mon Mar 17, 2014 1:47 pm

Post by addendum » Mon Jul 24, 2017 2:36 pm

Second what Dave said. Just curious, where did you read or hear that the output range was variable? Even the drawings on the module (at least the original Buchla faceplate - don't know what people might be using these days to avoid the controversy, or not), they show clearly that it's about distribution, not range.

User avatar
njdoyle
Common Wiggler
Posts: 171
Joined: Tue Jan 28, 2014 5:00 pm
Location: Ottawa
Contact:

Post by njdoyle » Mon Jul 24, 2017 3:18 pm

So to my second point about ranges, I've had a discussion off thread and the ranges are what are to be expected from this module and make sense in a Buchla mindset. I find it difficult to find definitive documentation for much anything Buchla and figured the design of the Doepefer output ranges made so much sense that it would be reflected in the original design. It makes sense from a western scale approach at least so I guess that's why it might make sense for it not to be that way in Buchla. :)

As for distributions, I did look at your site, Dave, to try and decipher things. From your scope image, I get the cyan signal from the 2^n output and the magenta signal from the n+1 output. This is still backwards from what I expect. This backwardness is consistent with the Sputnik West Coast Random Source though.

Image

User avatar
davebr
Veteran Wiggler
Posts: 668
Joined: Mon Sep 07, 2009 12:55 am
Location: Portland, OR USA

Post by davebr » Mon Jul 24, 2017 3:43 pm

njdoyle wrote:As for distributions, I did look at your site, Dave, to try and decipher things. From your scope image, I get the cyan signal from the 2^n output and the magenta signal from the n+1 output. This is still backwards from what I expect. This backwardness is consistent with the Sputnik West Coast Random Source though.

Image
From the manual ...
Image

User avatar
addendum
Common Wiggler
Posts: 127
Joined: Mon Mar 17, 2014 1:47 pm

Post by addendum » Mon Jul 24, 2017 4:37 pm

Good to know Doepfer's implementation differs from Buchla's. I think they're both valid and come from solid background philosophies. Personally I find the variable distribution more musically useful and exciting than variable range. You can always run the output through a scale quantizer. And in the Buchla systems you can create variable range yourself by running the output through a CV processor. Use the same modulation that you wanted to use on the 266' non-existing range parameter to attenuate the 266 output, same as you would do with a DC coupled VCA in a non-separating system.

User avatar
Cobramatic
Wiggling with Experience
Posts: 464
Joined: Sat Sep 21, 2013 6:45 pm
Location: Sydney

Post by Cobramatic » Mon Jul 24, 2017 7:26 pm

On a related question for this module - can somebody tell me exactly with the "integrator' does?
I have the 266e which does not have this function so I've never had the chance to experience what I might use it for.

I like the fact that the original 266 also has the S&H (or rather track and hold) and the alternative Pulse outs too.

User avatar
addendum
Common Wiggler
Posts: 127
Joined: Mon Mar 17, 2014 1:47 pm

Post by addendum » Mon Jul 24, 2017 7:36 pm

Integrator = slew limiter = lag processor. Agreed, every system should have an S&H/ T&H. The 266 integrator isn't optimal tho as it has only one control for rising and falling slopes.

User avatar
Cobramatic
Wiggling with Experience
Posts: 464
Joined: Sat Sep 21, 2013 6:45 pm
Location: Sydney

Post by Cobramatic » Mon Jul 24, 2017 7:42 pm

addendum wrote:Integrator = slew limiter = lag processor. Agreed, every system should have an S&H/ T&H. The 266 integrator isn't optimal tho as it has only one control for rising and falling slopes.
Thanks - that makes sense.
I have the 257e for that function - much more control (and 6 of them!)

User avatar
muncky
Ultra Wiggler
Posts: 801
Joined: Thu Sep 09, 2010 9:01 am
Location: Brighton
Contact:

Post by muncky » Tue Jul 25, 2017 1:17 am

Just found another series of super helpful Todd vids. Here's the 266 one:

User avatar
njdoyle
Common Wiggler
Posts: 171
Joined: Tue Jan 28, 2014 5:00 pm
Location: Ottawa
Contact:

Post by njdoyle » Tue Jul 25, 2017 8:42 am

I guess what I'm getting at is this:

In the scope image, the cyan line is an n+1 distribution. You can see that it would have seven values (if the few lower values that didn't occur are accounted for). You can also see that it generally sticks to the middle of the range. I am getting this signal out of the output marked as n^2.

The magenta line in the scope image clearly has more than seven values and is distributed pretty evenly over the entire range. It is a n^2 distribution. I am getting this signal out of the output marked as n+1.


It seems clear to me that, on my 266r, the outputs for QRV are the reverse as they are labeled. I have also noticed that on the Sputnik West Coast Random Source, this reversal is also present. The Doepfer version does not have it's distributions mislabeled.

The behaviour of my QRV outputs are actually opposite of Todd's in the video muncky linked.

Post Reply

Return to “Buchla, EMS & Serge”